IAEA-TECDOC-1661
Mitigation of Hydrogen Hazards in Severe Accidents in Nuclear Power Plants
INTERNATIONAL ATOMIC ENERGY AGENCY VIENNA
ISBN 978–92–0–116510–7 ISSN 1011–4289
IAEA-TECDOC-1661 nMITIgATIOn Of HyDrOgEn HAzArDs In sEvErE ACCIDEnTs In nuClEAr POwEr PlAnTs
IAEA SAFETY RELATED PUBLICATIONS
IAEA SAFETY STANDARDS
Under the terms of Article III of its Statute, the IAEA is authorized to establish or adopt standards of safety for protection of health and minimization of danger to life and property, and to provide for the application of these standards.
The publications by means of which the IAEA establishes standards are issued in the IAEA Safety Standards Series. This series covers nuclear safety, radiation safety, transport safety and waste safety. The publication categories in the series are Safety Fundamentals, Safety Requirements and Safety Guides.
Information on the IAEA’s safety standards programme is available at the IAEA Internet site
http://www-ns.iaea.org/standards/
The site provides the texts in English of published and draft safety standards. The texts of safety standards issued in Arabic, Chinese, French, Russian and Spanish, the IAEA Safety Glossary and a status report for safety standards under development are also available. For further information, please contact the IAEA at PO Box 100, 1400 Vienna, Austria.
All users of IAEA safety standards are invited to inform the IAEA of experience in their use (e.g. as a basis for national regulations, for safety reviews and for training courses) for the purpose of ensuring that they continue to meet users’ needs.
Information may be provided via the IAEA Internet site or by post, as above, or by email to [email protected].
OTHER SAFETY RELATED PUBLICATIONS
The IAEA provides for the application of the standards and, under the terms of Articles III and VIII.C of its Statute, makes available and fosters the exchange of information relating to peaceful nuclear activities and serves as an intermediary among its Member States for this purpose.
Reports on safety and protection in nuclear activities are issued as Safety Reports, which provide practical examples and detailed methods that can be used in support of the safety standards.
Other safety related IAEA publications are issued as Radiological Assessment Reports, the International Nuclear Safety Group’s INSAG Reports, Technical Reports and TECDOCs. The IAEA also issues reports on radiological accidents, training manuals and practical manuals, and other special safety related publications. Security related publications are issued in the IAEA Nuclear Security Series.
Mitigation of Hydrogen Hazards in Severe Accidents
in Nuclear Power Plants
AFGHANISTAN ALBANIA ALGERIA ANGOLA ARGENTINA ARMENIA AUSTRALIA AUSTRIA AZERBAIJAN BAHRAIN BANGLADESH BELARUS BELGIUM BELIZE BENIN BOLIVIA
BOSNIA AND HERZEGOVINA BOTSWANA
BRAZIL BULGARIA BURKINA FASO BURUNDI CAMBODIA CAMEROON CANADA
CENTRAL AFRICAN REPUBLIC CHAD CHILE CHINA COLOMBIA CONGO COSTA RICA CÔTE D’IVOIRE CROATIA CUBA CYPRUS
CZECH REPUBLIC DEMOCRATIC REPUBLIC OF THE CONGO DENMARK
DOMINICAN REPUBLIC ECUADOR
EGYPT EL SALVADOR ERITREA ESTONIA ETHIOPIA FINLAND FRANCE GABON GEORGIA GERMANY
GHANA GREECE GUATEMALA HAITI HOLY SEE HONDURAS HUNGARY ICELAND INDIA INDONESIA
IRAN, ISLAMIC REPUBLIC OF IRAQ
IRELAND ISRAEL ITALY JAMAICA JAPAN JORDAN KAZAKHSTAN KENYA
KOREA, REPUBLIC OF KUWAIT
KYRGYZSTAN LATVIA LEBANON LESOTHO LIBERIA
LIBYAN ARAB JAMAHIRIYA LIECHTENSTEIN
LITHUANIA LUXEMBOURG MADAGASCAR MALAWI MALAYSIA MALI MALTA
MARSHALL ISLANDS MAURITANIA MAURITIUS MEXICO MONACO MONGOLIA MONTENEGRO MOROCCO MOZAMBIQUE MYANMAR NAMIBIA NEPAL
NETHERLANDS NEW ZEALAND NICARAGUA NIGER NIGERIA
NORWAY OMAN PAKISTAN PALAU PANAMA PARAGUAY PERU PHILIPPINES POLAND PORTUGAL QATAR
REPUBLIC OF MOLDOVA ROMANIA
RUSSIAN FEDERATION SAUDI ARABIA SENEGAL SERBIA SEYCHELLES SIERRA LEONE SINGAPORE SLOVAKIA SLOVENIA SOUTH AFRICA SPAIN
SRI LANKA SUDAN SWEDEN SWITZERLAND
SYRIAN ARAB REPUBLIC TAJIKISTAN
THAILAND
THE FORMER YUGOSLAV REPUBLIC OF MACEDONIA TUNISIA
TURKEY UGANDA UKRAINE
UNITED ARAB EMIRATES UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND UNITED REPUBLIC OF TANZANIA
UNITED STATES OF AMERICA URUGUAY
UZBEKISTAN VENEZUELA VIETNAM YEMEN ZAMBIA ZIMBABWE
The Agency’s Statute was approved on 23 October 1956 by the Conference on the Statute of the IAEA held at United Nations The following States are Members of the International Atomic Energy Agency:
IAEA-TECDOC-1661
MITIGATION OF HYDROGEN
HAZARDS IN SEVERE ACCIDENTS IN NUCLEAR POWER PLANTS
INTERNATIONAL ATOMIC ENERGY AGENCY
VIENNA, 2011
COPYRIGHT NOTICE
All IAEA scientific and technical publications are protected by the terms of the Universal Copyright Convention as adopted in 1952 (Berne) and as revised in 1972 (Paris). The copyright has since been extended by the World Intellectual Property Organization (Geneva) to include electronic and virtual intellectual property. Permission to use whole or parts of texts contained in IAEA publications in printed or electronic form must be obtained and is usually subject to royalty agreements. Proposals for non-commercial reproductions and translations are welcomed and considered on a case-by-case basis. Enquiries should be addressed to the IAEA Publishing Section at:
Sales and Promotion, Publishing Section International Atomic Energy Agency Vienna International Centre
PO Box 100
1400 Vienna, Austria fax: +43 1 2600 29302 tel.: +43 1 2600 22417
email: [email protected] http://www.iaea.org/books
For further information on this publication, please contact:
Safety Assessment Section International Atomic Energy Agency
Vienna International Centre PO Box 100
1400 Vienna, Austria email: [email protected]
MITIGATION OF HYDROGEN HAZARDS IN SEVERE ACCIDENTS IN NUCLEAR POWER PLANTS
IAEA, VIENNA, 2011 IAEA-TECDOC-1661 ISBN 978-92-0-116510-7
ISSN 1011-4289
© IAEA, 2011
Printed by the IAEA in Austria
FOREWORD
Consideration of severe accidents in nuclear power plants is an essential component of the defence in depth approach in nuclear safety. Severe accidents have very low probabilities of occurring, but may have significant consequences resulting from the degradation of nuclear fuel.
The generation of hydrogen and the risk of hydrogen combustion, as well as other phenomena leading to overpressurization of the reactor containment in case of severe accidents, represent complex safety issues in relation to accident management. The combustion of hydrogen, produced primarily as a result of heated zirconium metal reacting with steam, can create short term overpressure or detonation forces that may exceed the strength of the containment structure. An understanding of these phenomena is crucial for planning and implementing effective accident management measures. Analysis of all the issues relating to hydrogen risk is an important step for any measure that is aimed at the prevention or mitigation of hydrogen combustion in reactor containments.
The main objective of this publication is to contribute to the implementation of IAEA Safety Standards, in particular, two IAEA Safety Requirements: Safety of Nuclear Power Plants: Design and Safety of Nuclear Power Plants: Operation. These Requirements publications discuss computational analysis of severe accidents and accident management programmes in nuclear power plants. Specifically with regard to the risk posed by hydrogen in nuclear power reactors, computational analysis of severe accidents considers hydrogen sources, hydrogen distribution, hydrogen combustion and control and mitigation measures for hydrogen, while accident management programmes are aimed at mitigating hydrogen hazards in reactor containments.
The IAEA staff member responsible for this publication was C.O. Park of the Division of Nuclear Installation Safety.
EDITORIAL NOTE
The use of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries.
The mention of names of specific companies or products (whether or not indicated as registered) does not imply any intention to infringe proprietary rights, nor should it be construed as an endorsement
CONTENTS
1. INTRODUCTION ... 1
1.1.BACKGROUND ... 1
1.2.OBJECTIVE AND SCOPE ... 5
1.3.STRUCTURE ... 5
2. POTENTIAL HYDROGEN SOURCES DURING THE EVOLUTION OF A SEVERE ACCIDENT ... 6
2.1.INTRODUCTION ... 6
2.2.IN-VESSEL HYDROGEN SOURCE ... 7
2.2.1. Short description of core degradation during a severe accident ... 7
2.2.2. In-vessel hydrogen source from Zr oxidation... 9
2.2 3. In-vessel hydrogen production coming from steel oxidation ... 14
2.2.4. In-vessel hydrogen production coming from B4C absorber material oxidation ... 14
2.2.5. Consequences to be drawn regarding calculations ... 16
2.3.EX-VESSEL HYDROGEN PRODUCTION ... 16
2.3.1. Short term H2 release during vessel lower head failure ... 16
2.3.2. H2 production during molten core-concrete interaction ... 17
2.3.3. Other possible ex-vessel H2 production ... 19
3. HYDROGEN DISTRIBUTION ... 19
3.1.DESCRIPTION OF CONTAINMENT ... 20
3.1.1. Full pressure containment ... 20
3.1.2. Containments with pressure suppression system ... 21
3.2.LOCATION OF HYDROGEN SOURCES IN THE CONTAINMENT ... 23
3.3.EFFECT OF RELEASE MODE AND SPRAYING ON HYDROGEN DISTRIBUTION ... 23
3.4.CONTAINMENT LAYOUT EFFECTS ... 24
3.5.ANALYTICAL TOOLS ... 25
3.5.1. Integrated codes or system codes ... 25
3.5.2. Lumped parameter codes ... 25
3.5.3. Computational fluid dynamics codes ... 26
3.5.4. Hybrid codes ... 28
3.5.5. Comparison of general advantages and disadvantages of the different code types. 29 3.6.EXPERIMENTAL FACILITIES TO MEASURE HYDROGEN DISTRIBUTIONS ... 29
3.6.1. Gas distribution experiments for large dry containments ... 29
3.6.2. Experiments for ice condenser containments ... 31
3.6.3. Recent and future experiments ... 31
4. HYDROGEN COMBUSTION ... 33
4.1.INTRODUCTION ... 33
4.2.FLAMMABILITY AND IGNITION CONDITIONS ... 34
4.2.1. Flammability ... 34
4.2.2. Auto ignition and ignition ... 35
4.3.MODES OF COMBUSTION ... 36
4.3.1. Deflagration ... 36
4.3.2. Detonation ... 36
4.3.3. Flame acceleration and deflagration to-detonation transition ... 38
4.3.4. Quenching ... 41
4.3.5. Mechanisms involved in Deflagration to Detonation Transition ... 41
4.3.6. Necessary criteria for flame acceleration and DDT ... 41
4.3.7. Diffusion flames ... 46
4.3.8. Effect of carbon monoxide ... 47
4.3.9. Pressure loads associated with different combustion phenomena ... 48
4.4.ANALYTICAL TOOLS ... 49
4.4.1. Combustion in containment systems codes (lumped parameter) ... 49
4.4.2. Combustion in CFD codes ... 50
4.5.EXPERIMENTAL FACILITIES ... 54
4.5.1. Small scale facilities ... 54
4.5.2. Medium test facilities ... 56
4.5.3. Experiments in large scale and complex geometries ... 56
5. RISK FROM HYDROGEN COMBUSTION... 58
5.1.COMBUSTION LOADS AND STRUCTURAL RESPONSE ... 58
5.2.THREATS FROM COMBUSTION TO THE CONTAINMENT ... 60
5.2.1. Direct damage. ... 60
5.2.2. Indirect damage ... 61
5.2.3. Failure of secondary containment ... 62
5.2.4. Failure of containment vent. ... 62
5.2.5. Effect of temperature. ... 63
5.3.SCENARIO EFFECTS ... 63
5.4.OTHER FACTORS RELEVANT FOR THE RISK FROM COMBUSTION GASES ... 64
5.4.1. Other substances than hydrogen. ... 64
5.4.2. Pressure build-up by non-condensables. ... 64
5.4.3. Stratification of gases ... 65
5.4.4. Effects from accident management ... 65
5.5.SENSITIVITY OF VARIOUS CONTAINMENTS TO HYDROGEN COMBUSTION LOADS ... 66
5.5.1. Large dry containment ... 67
5.5.2. Ice condenser containment ... 67
5.5.3. Suppression pool containment (BWR) ... 67
5.5.4. WWER confinement... 67
5.5.5. Future designs ... 68
5.6.TREATMENT OF HYDROGEN IN THE PSA ... 70
6. HYDROGEN MEASUREMENT ... 71
6.1.OBJECTIVES OF A H2 MEASUREMENT SYSTEM ... 71
6.2.H2 MEASUREMENT SYSTEMS ... 71
6.2.1. Hydrogen concentration measurement system inside the containment ... 72
6.2.2. Systems based on sampling ... 72
6.3.H2 MEASUREMENT SYSTEMS QUALIFICATION ... 73
6.4.H2 MEASUREMENT PROBE POSITION ... 73
6.5.H2 MEASUREMENT EVALUATION WITHOUT ANY H2 MEASUREMENT ... 73
7. HYDROGEN CONTROL AND RISK MITIGATION ... 73
7.1.INERTIZATION OF THE CONTAINMENT ATMOSPHERE ... 74
7.2.POST-ACCIDENT DILUTION ... 75
7.3.EARLY VENTING ... 75
7.4.HYDROGEN REMOVAL ... 75
7.4.1. Deliberate ignition ... 75
7.4.2. Spontaneous ignition ... 77
7.4.3. Catalytic recombination ... 77
7.5.STRATEGIC COMBINATIONS ... 77
7.5.1. Catalytic recombiners and igniters (dual concept) ... 77
7.5.2. Catalytic recombination and post-CO2 Injection ... 78
7.6.HYDROGEN CONTROL IN SEVERE ACCIDENT MANAGEMENT GUIDE APPROACHES ... 78
8. ANALYTICAL ASPECTS FOR HYDROGEN ANALYSIS ... 80
8.1.INTRODUCTION ... 80
8.2.DEFINITION OF THE PROBLEM ... 81
8.3.CONSERVATIVE VERSUS BEST-ESTIMATE CALCULATION ... 83
8.4.CHOICE OF COMPUTER CODE ... 83
8.5.CHOICE OF THE ACCIDENT SCENARIO ... 85
8.6.MODELLING FOR THE MITIGATION SYSTEM FOR HYDROGEN ... 86
8.7.HYDROGEN AND STEAM SOURCES ... 86
8.8.HYDROGEN DISTRIBUTION ... 86
8.8.1. Lumped parameter approach ... 87
8.8.2. CFD approach ... 89
8.9.IGNITION MODELLING ... 89
8.10.COMBUSTION MODELLING ... 90
8.10.1. Flame acceleration and DDT criteria applications ... 90
8.10.2. General data for combustion analysis ... 91
8.10.3. Combustion modelling in lumped parameter approach ... 92
8.10.4. Modelling of turbulence and combustion interactions in the CFD approach ... 92
8.10.5. DDT modelling ... 94
8.10.6. Detonation modelling ... 95
8.11.NUMERICAL ISSUES ... 96
8.11.1. Convergence ... 96
8.11.2. Stability ... 96
8.11.3. CPU time ... 97
8.12.TREATMENT OF UNCERTAINTIES ... 97
8.12.1. Modelling uncertainties ... 98
8.12.2. User effect ... 100
8.12.3. Use of sensitivity analysis ... 101
8.13.BENCHMARKING ... 102
8.14.EFFECT OF SCALING ... 104
9. REMAINING ISSUES IN HYDROGEN RISK MITIGATION ... 105
9.1.INTRODUCTION ... 105
9.2.REMAINING ISSUES ON HYDROGEN AMOUNT AND PRODUCTION RATE ... 106
9.3.REMAINING ISSUES ON HYDROGEN DISTRIBUTION ... 106
9.4.REMAINING ISSUES ON HYDROGEN COMBUSTION ... 108
9.5.REMAINING ISSUES ON ANALYTICAL ASPECTS ... 109
REFERENCES ... 119
ANNEX I. EXPERIMENTAL FACILITIES TO INVESTIGATE
HYDROGEN SOURCE ... 125
ANNEX II EXPERIMENTAL FACILITIES TO MEASURE HYDROGEN DISTRIBUTION ... 137
ANNEX III COMBUSTION EXPERIMENT FACILITIES ... 149
ABBREVIATIONS ... 155
CONTRIBUTORS TO DRAFTING AND REVIEW ... 157
1. INTRODUCTION 1.1. Background
The basic goal of severe accident management in nuclear power plants (NPPs) is the protection of the containment integrity and the containment becomes the ultimate barrier against the release of fission products to the environment. There are various potential challenges to the containment integrity during a severe accident in a light water reactor (LWR). The combustion of hydrogen, produced primarily as a result of heated zirconium metal reacting with steam, can create short term pressure or detonation forces that may exceed the strength of the containment structure and lead to early containment failure. For most NPPs, severe accidents lead to hydrogen release rates that exceed the capacity of hydrogen control measures at conventional design basis accident (DBA). Local high hydrogen concentrations can be reached in a short time, leading to combustible gas mixtures in the containment. Moreover, a long term pressure build-up may occur due to steam generation through decay heat and/or through the generation of non-condensable gas from the interaction of the molten core with the containment basemat concrete.
Hydrogen production, distribution and combustion in post-accident containment are very complex and highly plant- and scenario-specific phenomena. Hydrogen combustion can take place in a variety of forms: mild deflagration, fast or accelerated flames, deflagration to- detonation transition (DDT) and detonation. In order to study the influence and mitigate the consequences of hydrogen combustion, detailed studies are needed to determine the hydrogen generation rates and overall amount released to the containment. The distribution of the hydrogen released within the containment determines local and global hydrogen concentrations, which are decisive for the evaluation of the various combustion modes, such as diffusion flames, deflagration and detonation, depending on geometrical effects and concentrations. An understanding of all these phenomena is crucial for planning and implementing effective hydrogen management measures. These measures include enhancement of mixing, deliberate combustion through igniters, use of recombiners, and post-accident inerting.
In most countries, there are no strict regulatory requirements on the implementation of hydrogen mitigation measures for existing plants, while for reactors that are planned or under construction, these measures have to be incorporated into the design. Therefore, mitigation measures for hydrogen already exist in some countries. Nonetheless, these measures have been implemented in many NPPs on a voluntary basis. The need for such measures was derived both from deterministic consideration of plant vulnerabilities and from risk investigations, but sometimes also owing to non-technical reasons. The implementation level varies significantly from country to country and even from plant to plant in an individual country. In some plants, no decision has been taken yet on the implementation of measures. In other plants, measures have been implemented to cope with hydrogen produced during DBAs.
In many plants, the measures are already capable to cope with hydrogen produced during severe accidents. In some cases, a final decision on this partial issue has been postponed after specification of an overall approach to severe accident management. The issue of mitigation measures for hydrogen is in progress in a number of countries. It is therefore appropriate to share views and experiences among the countries.
Since combustion of hydrogen represents a severe challenge to the containment integrity, mitigation measures for hydrogen are one of the essential parts of any accident
management programme and all the IAEA publications developed for accident management are also applicable to the hydrogen issues.
According to the IAEA INSAG-10 [1], consideration of highly improbable severe plant conditions is an important component of defence in depth and measures aimed at controlling the course of severe accidents and mitigation of their consequences need to be available. The most important objective of accident management is the protection of the containment.
According to this publication, management of severe accidents has to be flexible enough to take into account many uncertainties about the actual course of a severe accident. The importance of adequate instrumentation qualified for accident conditions is also stated.
Among basic safety principles specified in IAEA INSAG-12 [2] there are three principles specifically devoted to management of accidents beyond the design basis: the need for development of strategies for accident management, training and procedures for accident management and engineered features for accident management.
The IAEA has developed a number of publications that provide guidance and support in severe accident analysis and accident management at NPPs [3–10]. Various strategies are discussed in Ref. [11], which provides a comprehensive overview and comparison of mitigation measures for hydrogen in severe accidents. However, this publication is limited in scope and requires updating, as the information it contains represents the level of knowledge of the early 1990s.
Several comprehensive publications dealing with the hydrogen issue have been also developed within the framework of OECD/NEA activities [12–14]. All these publications represent a valuable source of information for any further work on hydrogen issues.
Consideration of severe accidents and accident management in plant design and operation is stipulated by the IAEA Safety Requirements publications Safety of Nuclear Power Plants:
Design [3] and Safety of Nuclear Power Plants: Operation [4], respectively. According to these requirements, control of severe accidents is an important part of defence in depth. Among other plant design requirements it is also stated in Ref. [3] that:
“5.31. Certain very low probability plant states that are beyond design basis accident conditions and which may arise owing to multiple failures of safety systems leading to significant core degradation may jeopardize the integrity of many or all of the barriers to the release of radioactive material. These event sequences are called severe accidents.
Consideration shall be given to these severe accident sequences, using a combination of engineering judgement and probabilistic methods, to determine those sequences for which reasonably practicable preventive or mitigatory measures can be identified.
Acceptable measures need not involve the application of conservative engineering practices used in setting and evaluating design basis accidents, but rather should be based upon realistic or best estimate assumptions, methods and analytical criteria. On the basis of operational experience, relevant safety analysis and results from safety research, design activities for addressing severe accidents shall take into account the following (Ref. [3] paras 531, 531 (1, 3, 6):
(1) Important event sequences that may lead to a severe accident shall be identified using a combination of probabilistic methods, deterministic methods and sound engineering judgement.
(3) Potential design changes or procedural changes that could either reduce the likelihood of these selected events, or mitigate their consequences if these selected events were to occur, shall be evaluated and shall be implemented if reasonably practicable.
(6) Accident management procedures shall be established, taking into account representative and dominant severe accident scenarios.”
Ref. [4] establishes the following requirements on severe accidents and accident management in the operation of nuclear power plants:
“3.12 Plant staff shall receive instructions in the management of accidents beyond the design basis. The training of operating personnel shall ensure their familiarity with the symptoms of accidents beyond the design basis and with the procedures for accident management.
5.12 (…) Emergency operating procedures or guidance for managing severe accidents (beyond the design basis) shall be developed.”
Further details on design considerations for severe accidents can be found in the IAEA Safety Guide on Design of Reactor Containment Systems for Nuclear Power Plants [15]
which makes distinction between existing and new plants in accident management. For existing plants, severe accidents are carefully analysed in order to identify safety margins and accident management measures. For new plants, consideration of severe accidents are aimed at practically eliminating the damage to the containment in both the early and the late phase, severe accident conditions with an open containment or with containment by-pass. In addition, the publication provides more specific guidance related to capability of containment systems under severe accident conditions, including:
• Making available proper instrumentation and procedures to initiate preventive or mitigatory measures;
• Verifying the necessary survivability of equipment and instrumentation under severe accident conditions;
• Ensuring integrity and leaktightness of the containment; for existing plants “as far as this can be achieved by reasonable means”;
• Preventing combustion or deflagration of hydrogen potentially damaging the containment systems;
• Providing for hydrogen and other combustible gases monitoring as well as for monitoring of other parameters important for performing of severe accident management guidelines.
In addition to the aforementioned Standards established from a design perspective, there are four IAEA Safety Standards [5, 98, 99, 100] that deal with in part how to perform safety analysis including severe accident analysis. Reference [5] specifies the generally applicable requirements to be fulfilled in safety analysis for all facilities and activities relevant to radiation risks. It is stated in the Ref. [5] that the consequences associated with beyond design basis accidents (BDBAs) have to be addressed in the safety analysis. Also stated is that both deterministic and probabilistic approache have to be included in the safety analysis. Reference
[98] provides detailed guidance on deterministic safety analysis whereas Ref. [99] and Ref.
[100] describe guidance on level 1 and level 2 probabilistic safety assessments for nuclear power plants, respectively.
There are also three Safety Reports [7–9] and several other IAEA safety related publications dealing with severe accidents. The IAEA Safety Report on Accident Analysis of Nuclear Power Plants [8] gives a practical general guidance for performing accident analysis based on present good practice worldwide. The report concentrates on analysis of accidents within the design basis, but includes also some suggestions for analysis of severe accidents. It describes basic approaches used for severe accident analysis, characterization of initiating events and scenarios to be analysed, designators for categorization of accident sequences, overview of recovery strategies for prevention and mitigation of severe core damage, basic characteristics of computer codes and acceptance criteria used in severe accidents domain, applicability of results of analysis. More specific information on methodology for severe accident analysis can be found in the IAEA Safety Report on Approaches and Tools for Severe Accident Analysis for Nuclear Power Plants [9]. The publications include description and status in modelling of severe accident phenomena, examples of acceptance criteria for severe accident domain, overview of applicable computer codes, their characteristics and status of validation, approaches recommended for various applications of analysis and advice on selection of codes, advice on selection of scenarios, consideration of essential design characteristics, basic steps in developing data and calculations, main suggestions for best estimate calculations, presentation of results, and consideration of uncertainties. Issues related to the hydrogen are also addressed, but due to the broad scope of the publications, hydrogen specific information is very limited.
Detailed information on approaches applicable for development of accident management programmes has been published in the IAEA Safety Report on Implementation of Accident Management Programmes in Nuclear Power Plants [10], which is based on developments that have been made in the field of accident management worldwide. The Safety Report provides a description of all elements that need to be addressed by the team responsible for the preparation, development and implementation of a plant specific accident management programme at a NPP. The elements include the establishment of the team, selection of accident management strategies, safety analyses required, evaluation of the plant systems performance, development of accident management procedures and guidelines, staffing and qualification of accident management personnel, and training needs. However, the Safety Report does not contain any technical details regarding mitigation measures for hydrogen in severe accidents.
The publication on the approaches to the safety of future NPPs [16] also specifies that severe accidents have to be considered explicitly in the design, but as a separate category.
Systems added to the design to address severe accidents need to be of high quality, but safety grade quality levels are usually not required. Particular attention is devoted to the preservation of the containment integrity and leaktightness. For analysis of severe accident sequences selected and addressed in the design, best estimate assumptions, methods and data are the preferred tools to avoid distortion of the physical picture. Accident management is supported by NPP design features, aimed at providing time for management actions and to contributing to smooth plant response and to simplification of emergency planning.
The need for development of specific guidance to cover major issues relating to hydrogen sources, hydrogen distribution, hydrogen combustion, prevention of hydrogen combustion, control and mitigation measures for hydrogen was extensively discussed at the IAEA Technical Committee Meeting on Implementation of Hydrogen Mitigation Techniques
and Filtered Containment Venting [17]. Participants at this meeting felt that an up to date IAEA publication on mitigation measures for hydrogen was needed, provided that the publication would concentrate on the practical aspects of the implementation of the measures for various reactor designs, with reference to already existing publications.
1.2. Objective and scope
The main objective of the present publication is to contribute to the implementation of relevant IAEA Safety Standards, in particular regarding two requirements (Refs [3, 4]):
• Performing computational analysis of severe accidents, and notably all problems related to hydrogen sources, hydrogen distribution, hydrogen combustion, hydrogen control and mitigation measures;
• Development and implementation of accident management programmes in NPPs, notably of those measures which are aimed at mitigation of hydrogen in the reactor containments.
This publication is intended as a self-standing report, it is suggested for the user to read also other relevant guidance publications to learn more comprehensively about accident management. This publication is aimed to be useful for utilities, as well as for regulatory bodies and their technical support organizations.
Although this publication does not explicitly differentiate among various reactor types, it has been written essentially on the basis of available knowledge and databases developed for LWRs. Therefore, its application is mostly oriented towards PWRs (including WWERs) and BWRs. However, it can be also used as a preliminary publication for other types of reactors such as PHWRs and RBMKs. For the reasons stated above, this publication is more appropriate for existing NPPs, although to large extent it contains information which is also applicable to new reactor designs. Since hydrogen mitigation measures are typically plant specific, the practical solutions for different reactor designs in the present publication should be considered as examples that are not intended to be adopted without a critical evaluation.
1.3. Structure
The present publication consists of eight sections. Section 2 discusses potential hydrogen sources during a severe accident. They are described at various stages, such as from in-core degradation to core-concrete interaction. Sections 3 and 4 describe the distribution and combustion of hydrogen, including calculation tools and related experimental facilities. These sections consider the effects of release mode and spraying on hydrogen distribution, containment layout effects and combustion modes such as deflagration, detonation and flame acceleration. Section 5 discusses possible risks from hydrogen combustion and the evaluation of risk due to static and dynamic loads on the containment. Section 6 reviews hydrogen measurement systems, systems qualification and probe position. Section 7 introduces various techniques for the control of hydrogen and mitigation measures such as inertization, post-accident dilution, early venting and hydrogen removal. Section 8 considers analytical aspects for hydrogen behaviour analysis. Also discussed in this Section are calculation methodology, computer codes, scenarios, models and other numerical aspects. Section 9 reviews the remaining issues in mitigation measures for hydrogen. Finally, Annexes I to IV describe detailed experimental facilities to investigate hydrogen sources, to measure hydrogen distribution and combustion and practical examples of techniques for mitigating hydrogen hazards.
5
2. POTENTIAL HYDROGEN SOURCES
DURING THE EVOLUTION OF A SEVERE ACCIDENT 2.1. Introduction
Potential hydrogen sources during the development of a severe accident in a LWR come from:
• In-vessel metal oxidation (Zr clads and grids and other metallic structures) or B4C absorber material oxidation with steam or with water contained in the reactor pressure vessel lower plenum,
• Ex-vessel oxidation of metallic material (Zr, Cr, Fe…) during direct containment heating (DCH) or into the water eventually contained in the cavity pit (short term event occurring at vessel lower head failure),
• Ex-vessel oxidation of metallic material (Zr, Cr, Fe…) during molten core concrete interaction (MCCI) (complete and rapid energetic oxidation of Zr and Cr during the first hour, then partial and slow oxidation of Fe up to the time of the complete basemat penetration by the corium).
Despite the fact that the water radiolysis (in-vessel or in the sump water of the containment) and the metal corrosion in the containment (mainly with Al and Zr) are taken into account in hydrogen sources during a DBA (e.g. loss of coolant accident), these sources are considered negligible during the development of a severe accident. As a matter of fact, in the event of a severe accident, several days are required before the amount of hydrogen produced by these two processes ‘alone’ renders the air flammable. For instance, typical figures for water radiolysis are around some hundreds of kg of hydrogen produced after three months and 100 kg of hydrogen produced by metal corrosion after several hours, i.e. far less than from other sources [14].
Boron carbide (B4C) is used as absorber material for BWRs, WWERs and some western PWRs. For instance, in the French P4 and P'4 type of PWRs, the upper part of the hybrid B4C/AIC control rods is composed of pellets contained inside a stainless steel cladding, contained in a Zr guide tube. Typical for PWR design, the B4C is in the form of sintered pellets, which have been sintered at atmospheric pressure. In the WWERs and BWRs, usually, the B4C is in the form of powder, contained inside a stainless steel cladding, and the control blade is also made of stainless steel. Table 1 gives some orders of magnitude of the B4C and Zr masses, for specific and typical PWRs, WWERS and BWRs.
Very rough orders of magnitude could be:
• 200 to 300 kg of B4C for PWRs with B4C compared to at least 4 times more for BWRs with the same power level,
• 20 to 30 tons of Zr for PWRs compared to at least 2 to 3 times more for BWRs with the same power level
TABLE 1. ORDERS OF MAGNITUDE OF THE B4C AND ZR MASSES, FOR SPECIFIC AND TYPICAL PWRS, WWERS AND BWRS
Typical PWR, kg
(3600 MW·th)
French P4-P'4 PWR,
kg (3800 MW·th)
French N4 PWR, kg
(4270 MW·th)
WWER-1000 Russian fuel,
kg [18]
WWER-1000 Westinghouse Fuel Temelin, Czech Rep., kg
[18]
Typical BWR, kg
(3800 MW·th)
[18]
B4C No B4C ~320 ~340 ~270 ~200 ~1200
Zr ~26000 ~28000 ~30000
~22630 (l% Nb cladding)
~24765 (with ~1090 kg
spacer grids)
~76000
UO2 ~105 ~1.15 × 105 ~1.24 × 105 ~80100 ~91750 ~1.55
× 105
2.2. In-vessel hydrogen source
2.2.1. Short description of core degradation during a severe accident Early core degradation
After the core is uncovered, the heat transfer from the fuel to the steam is small compared with decay heat produced and, hence, the fuel temperature increases. The high temperature leads to oxidation of the Zr fuel cladding and hydrogen generation and can also lead to clad ballooning and rupture. Clad rupture is the cause of the first fission product release. The heat-up rate can increase to well above 1 K/s as the local temperature increases above ~1300 K, due to rapid oxidation of Zr and the strongly exothermic nature of the reaction. This part of the core degradation with no loss of fuel rod-like geometry, and only metallic melt and blockage of some channels, is often considered as the ‘early phase of core degradation’.
During the core heat-up of this early degradation phase, hydrogen is mainly produced by steam oxidation of Zr cladding. For BWRs, B4C oxidation may also have some contribution.
For intact geometry, fresh metallic surfaces exposed at extreme temperatures oxidize, where the oxygen transport through the material is by gas diffusion. Otherwise, the solid state transport of oxygen determines the reaction and the growth rate of the oxidized layer.
From [14], it is commonly agreed that prediction of the hydrogen average source rate, without core reflooding, is typically about 0.2 kg/s for a typical 1000 MW(e) PWR, and this value is sufficiently accurate as long as the core geometry remains intact.
The Zr oxidation kinetics in a steam environment – for intact core geometry – is often described by diffusion models and parabolic correlations, based on experimental investigation, under the laws detailed in Table 2. For other fuel clad materials, such laws are defined in Table 2. Here δ is a kinetic constant under Arrhenius law.
TABLE 2. MAIN ZR OXIDATION CORRELATIONS WITH STEAM IN CLAD GEOMETRY
Correlation Details
Baker-Just [19]
δ = 4.059 exp(–190200/RT), δ in g2m–4s–1, T in K (1273<T<2123)
This correlation retains its importance for comparison and licensing purposes.
However, from a common understanding of EC 5th PCRD COLOSS project experts, it is advisable not to consider it in best-estimate calculations. It is commonly agreed to use the Baker-Just correlation for ‘envelope calculation’ type.
Urbanic-Heidrick [20]
δ = 0.036 exp(–139841/RT), δ in g2m–4s–1, T in K (1323<T<1853) δ = 0.108 exp(–138095/RT), δ in g2m–4s–1, T in K (1853<T<2123)
From EC 5th PCRD COLOSS project experts, doubts in the given specimen temperatures and their homogeneity, both related to the method of inductive heating, are fostered by strong discrepancies in comparison to data gained in test programmes using furnace heating. The low-temperature correlation over-estimates the oxidation towards lower temperature. The high-temperature branch under-estimates oxidation with increasing temperature.
Prater-Courtright [21]
δ = 0.3622 exp(–167200/RT), δ in g2m–4s–1, T in K (1783<T<2673), relative error ± 35%.
δ = 32.94 exp(–220000/RT), δ in g2m–4s–1, T in K (1573<T<1783), relative error ± 42%.
From EC 5th PCRD COLOSS project experts, in the temperature range above 1800 K, the Prater-Courtright correlations are the unrivalled only choice. Their precision is judged to be considerably inferior due to experimental procedures and evaluation methods, necessary for measuring fast reactions.
Erbacker-Leistikow
[22] δ = 0.524 exp(–174284/RT), δ in g2m–4s–1, T in K (1073<T<1783), relative error ± 42%.
Cathcart-Pawel [23]
δ = 0.362 exp(–167117/RT), δ in g2m–4s–1, T in K (1273<T<1573)
The Cathcart-Pawel and the Leistikow correlations are judged to be of equal and high reliability by EC 5th PCRD COLOSS project experts. This standard is understood to result from strong efforts towards precise temperature measurement and control, the volume of the data bases and adequate and consistent evaluation procedures. Similar results from other programmes confirm this judgment. For support of a choice between both sets of correlations, only comparably weak arguments might be mentioned, the more standardized procedures and the special temperature calibration efforts in favour of the Cathcart-Pawel correlations, the larger data base, the availability of experimentally determined mass gain (oxygen uptake) for all tests and the better fit for lower temperatures in favour of the Leistikow correlations.
Loss of core geometry
UO2 fuel can be liquefied at temperatures well below its melting point (3100 K) by dissolution in molten Zr or other metallic material such as iron. At higher temperatures, fuel liquefaction can occur due to the interaction between UO2 and ZrO2.
Because of their higher freezing temperature compared to metallic melts, the (U-Zr-O) melts created can lead to a channel blockage at a higher elevation in the core than the metallic blockage. As a result of the diversion of steam around the blockage and of the low thermal conductivity of this (U-Zr-O) crust, heat transfer inside the ceramic blockage is slow, and a molten pool can form inside the core within a ceramic crust.
During the loss of geometry of the core, experts presently consider that the main source of released H2 comes from (U-Zr-O) melt oxidation. Correct modelling of the (U-Zr-O) melt oxidation is still under investigation, based on newly dedicated experimental programmes.
Presently, in the severe accident core degradation codes, the Zr oxidation correlations established for intact clad geometry are often extrapolated well beyond their domain of validity, being also used to calculate the hydrogen production during the (U-Zr-O) melt oxidation and the reflooding of the damaged core during this phase of the degradation.
Consequently, a large uncertainty stills exists regarding the hydrogen production during this phase of the core degradation. Nevertheless, due to the decrease of metal surface in contact with steam during this phase, experts estimate that the H2 production can be lower than in the previous phase. But only the use of experimentally validated correlations in the code will confirm this assessment.
Late core phase degradation
Crust failure and melt relocation to the lower plenum are ‘late-phase’ core damage progression phenomena for which the uncertainties are greater than for the early phase phenomena. The general understanding of the late-phase phenomena is based on examination of the damaged TMI-2 core, because large scale experiments have not been run at high enough temperatures and for long enough time periods for the phenomena to occur fully.
In the late core degradation phase, hot melt can drop into the lower plenum of the vessel, which may be filled with water. Such injection of the melt into the water, for instance in the form of a jet, and fragmentation of the melt, would lead to an increase of the oxidation reaction surface (fragmented particles) and to a strong production of steam, which can oxidize the available metal. Experiments ZREX performed in the USA [24] with Zr/ZrO2 and Zr/stainless steel, with oxidation degrees of up to 40%, have indicated that typically 5 to 25%
of the metals are oxidized.
Presently, as mentioned above, in the severe accident core degradation codes, the Zr oxidation correlations established for intact clad geometry are often extrapolated well beyond their domain of validity, being also used to calculate the hydrogen production during the late reflooding of the damaged core and also during the oxidation of fragmented particles falling into water during this late core degradation phase.
Consequently, a large uncertainty stills exists regarding the hydrogen production during this phase of the core degradation.
2.2.2. In-vessel hydrogen source from Zr oxidation
Orders of magnitude of the H2 mass assuming a complete oxidation of 100% of the Zr mass with steam
Table 3 gives orders of magnitude of the H2 mass assuming a complete oxidation of 100% of the Zr mass with steam following the complete chemical reaction:
Zr + 2H2O → ZrO2 + 2 H2 + ∆H ∆H = - 586.6 kJ/mole Zr
Where ∆H is an energy released during the chemical reaction and 0.0442 kg H2 per Kg Zr was oxidized.
A very rough order of magnitude of hydrogen created by full Zr oxidation could be up to 1000 kg of H2 for a typical PWR compared to at least 3 to 4 time more for a BWR with the same power (around 1000 MW(e)), and around 1100 kg of H2 for a 1000 MW(e) WWER.
TABLE 3. ORDERS OF MAGNITUDE OF THE H2 MASS ASSUMING OXIDATION OF 100% ZR WITH STEAM FOR SPECIFIC AND TYPICAL PWRS, WWERS AND BWRS
Typical PWR, kg
(3600 MW·th)
WWER-1000 Russian fuel,
kg
WWER-1000 Westinghouse
fuel, kg
Typical BWR, kg (3800 MW·th)
Zr ~26000 ~22630 (l% Nb cladding)
~24765 (including 1090
kg spacer grids
~76000
H2 ~ 1150 ~1000 ~1095 ~3360
Main phenomena involved in the H2 release during a core degradation
The main physical phenomena controlling the in-vessel Zr oxidation are listed in the following sub-sections. For each phenomenon, the level of present knowledge or uncertainty is stated. From all these phenomena, the thermohydraulic effect is considered by experts to be the most influential.
1) Thermohydraulics
The Zr oxidation kinetics is strongly connected to the steam kinetics in the core. In a calculation, the behaviour of steam in the core depends on the thermohydraulic models of the code, which could be 1D (as in integrated severe accident codes, such as MELCOR, MAAP or ASTEC or 2D (in mechanistic core degradation codes, such as ICARE/CATHARE). But whatever the codes are, there exist two main events of the core thermohydrauylics which control the hydrogen release:
• The duration time of the core uncovery:
As long as steam starvation does not happen, the longer the uncovery time, the greater the amount of hydrogen produced by Zr oxidation, because the core has enough time to heat up and steam is always available.
Small break loss of coolant accident (LOCA) has a long core uncovery time and, as a consequence, its scenario must be included in the list of calculated scenarios.
• The redistribution of steam flow around a channel blockage:
During the early phase of the degradation, ballooning and deformation of the cladding leads to a redistribution of the steam in the core. Later on, the molten material relocation can plug some channels. These events prevent locally the arrival of steam on some available Zr surfaces and redistribute the steam in the surrounding core channels, leading to an increase of temperature in the core periphery, for instance.
The 1D or 2D thermohydraulic modelling approach of this event will give very different results in terms of hydrogen release, because of the very different flow redistribution scheme and associated heat transfer in the core. For example, very preliminary calculations using a
2D thermohydraulic model versus a 1D thermohydraulic model in the same code have been performed by IRSN with the ICARE/CATHARE code. Results show a large discrepancy between the calculated hydrogen releases, the 2D production giving the greater one.
Nevertheless, validation of such 2D flow redistribution models is difficult because of a lack of dedicated tests.
When using a 1D code and running calculations to check if no possible ‘cliff edge effect’ could exist in the plant regarding the hydrogen risk mitigation, the potential consequences of this lack of 2D modelling of the code need to be kept in mind.
2) Zr oxidation kinetic of the fuel rod clads (intact geometry)
The Zr oxidation kinetics is very often described by consistent diffusion models and parabolic correlations, based on experimental investigations, under the law:
k2 = δ t
where k = mass of oxidized Zr/m2, t = time in second, δ = kinetic constant under Arrhenius law as a function of temperature = Aexp(-Q/RT), Q= activation energy, R = 8.314 kJ/mol/K, T in Kelvin.
The codes using the Zr oxidation correlations for reactor calculations rely on the extrapolation of Zr oxidation correlations coming from specific tests, with given physical conditions.
Limits of the Zr oxidation correlations
Experiments used to obtain the Zr oxidation kinetics have some limits:
• All correlations have been obtained from tests using intact clad tube geometry:
They are used only during the so-called ‘early phase degradation’ phase of the core.
• Zr used in the tests was non-irradiated:
Presently, there exist no tests of Zr oxidation using irradiated Zr.
• The parabolic law used in the correlations has been established under isothermal tests:
The temperature gradient is less than 5 K/s to use the correlation for transient calculations post test calculations of temperature-transient experiments calculated have confirmed the use of these Zr correlations under those conditions; the reaction rate, however, can differ with time/temperature during the transient;..
• All tests have been done under atmospheric pressure (Annex I):
There exists presently only 2 tests done under elevated pressure, around 60 bar: the PBF-SFD-1.1 and the PBF-SFD-scoping test. These tests were in-pile core degradation tests. Some French calculations of these 2 tests have been performed using the ICARE-2 code (version V3 mod0.4) and the calculated H2 production was not under-estimated. But it would be hazardous to conclude anything from only 2 tests, because these tests had substantial uncertainties.
• These correlations require a correct modelling of the oxygen diffusivity around the clad.
Presently, there is no international consensus on the best Zr oxidation correlation to be used and the severe accident core degradation codes often offer the choice between several correlations.
• In addition to diffusion, there is oxygen transport through cracks and oxidation by spalling, as discussed above
For instance, the MAAP 4.04 uses the Cathcart-Pawel correlation for temperatures less than 1850 K and the Baker-Just correlation for higher temperatures and the ICARE-CATHARE (version V1Mod1.2) uses the Urbanic-Heidrick correlation.
Nevertheless, the use of some oxidation correlations, even those perfectly matching the experimental Zr oxidation results, can lead to largely overestimate the H2 production if e.g.
not a proper interface model is used to calculate the right amount of steam arriving at the Zr surface, such as the H2O diffusion in the (H2+H2O) mixture at the clad interface.
Consequently, most of the international severe accident codes presently use the Urbanic-Heidrick correlation because it tends to give better results when compared to integral tests such as PHEBUS, despite the present lack of correct modelling of the core thermal hydraulic in these codes (the diffusivity models of oxygen around the clad overestimate the oxygen diffusion).
When running calculations for a scenario, this effect needs to be kept in mind. If interface models are present in the code allowing to calculate the amount of steam arriving on the Zr surface, such as the H2O diffusion in the (H2+H2O) mixture at the clad interface, the Baker-Just correlation can e.g. be used for an ‘enveloping’ calculation, or the Cathcart-Pawel with Baker-Just (or Leistikow) correlation as in MAAP 4.04 for ‘best-estimate’ calculations.
If no such models are present, the Urbanic-Heidrick correlation still remains the usual choice.
3) Onset of ceramic melt relocation
Concerning the loss of core geometry, the onset of melt relocation and fuel rod collapse changes from one experimental result to another. This is partly caused by especially the chemical parameters, which are still not well understood today. The start of the loss of core geometry is relevant for the hydrogen production, because the loss of core geometry is linked to the decrease of the oxidation surface.
Any criterion used in a code linked to the onset of melt relocation plays a role in the calculated H2 release. E.g., in the French ICARE codes, the ‘clad failure criteria’ depend on temperature criteria and the thickness of ZrO2 created on the external surface of the clad. In the MAAP code, the approach is similar but there is in addition a simplified mechanistic failure criterion (creep failure criterion). Both codes have also chemical dissolution models of Zr with surrounding materials.
When running calculations for a scenario, this effect has to be kept in mind and run sensitivity cases, choosing the time of the onset of the ceramic melt relocation ‘later’ than the reference case choice. Help to find a proper value for this parameter needs to come either from the validation report of the used code or from the advised values list given in the code manual.
Choosing a parameter value allowing a ‘later core relocation’ increases the time of intact geometry rod for rod oxidation, which gives usually a higher hydrogen release, because
hydrogen releases of relocated core are judged to be lower than with an intact geometry, assuming the same parabolic oxidation correlation. This is due to the fact that that the metallic surface is decreased in a relocated core compared to an intact geometry. Note that the use of the oxidation correlation depends also on the possibility of the ‘break away’ effect, discussed before – this item is further treated in Item 4) below.
4) Oxidation of (U-O-Zr) melts
Some experimental work oxidation of (U-Zr-O) mixtures are in the frame of the COLOSS project of the EC-FWP-5, such as the SKODA-UJP experiments, using solid U-Zr-O alloys ingots (Annex I). Preliminary experimental results tend to show that Zr oxidation kinetics is linear, i.e. faster than the parabolic kinetic of the intact geometry Zr oxidation kinetics. This means that when melt relocation starts, the oxidation of U-Zr-O mixtures could contribute to a significantly larger hydrogen production, with a high kinetics.
Consequently, the use of parabolic Zr oxidation correlation (such as ones used for intact geometry) is not advised.
As parabolic correlations are presently used in codes and calculations have to be run to check if no possible ‘cliff edge effect’ exists in the plant regarding the hydrogen risk mitigation, the potential underestimation of the hydrogen source due to this lack of modelling of the (U-Zr-O) mixtures oxidation needs to be kept in mind.
5) In-vessel reflooding
Large uncertainties remain to exist concerning the hydrogen production in case of reflood of a degraded core. High hydrogen production rates associated with reflood have been estimated from the TMI accident and recorded in some integral experiments (CORA, LOFT, etc.). The current FZK QUENCH experiments have extended the knowledge base (see Annex I). But since this programme does not use prototypic materials and investigates mainly rod-like geometries, there will still be some remaining uncertainties.
Nevertheless, some quantitative and qualitative useful information can be presently extracted from all related experiments. IRSN conducted recently a synthesis of knowledge regarding in-vessel core-reflooding, based on analysis of international synthesis documents and re-analysis of some main experimental outcome [19]. The main conclusions of this work are:
• For a mostly intact core, it is rather unlikely to get a hydrogen peak during reflooding. In the QUENCH international workshops, consensus was achieved on a weak contribution of ZrO2 cracking and shattering (leading to oxidation of underlying Zr), and of clad hydriding on hydrogen release during reflooding. If
‘break away’ occurs, however, a large hydrogen source will occur due to the cracking and spallation, as was seen in the QUENCH-12 test for WWER-fuel [18].
• For a slightly degraded core, it is likely that the hydrogen peaks observed during reflooding in the tests come from the oxidation of metal rich (Zr-O, U-Zr-O) mixtures during their candling flow or after freezing. The H2 kinetics during reflooding in this core configuration is rather rapid (~ linear kinetic), compared to the parabolic kinetics of the oxidation of an intact Zr bundle. B4C seems to contribute to the H2 peak (QUENCH-07 and CORA tests).
• For a ‘particulate bed’ core, the oxidation of metal is responsible for significant additional H2 release, due to the increase of the exchange surface.
• Concerning the status of models and codes in 2002 to calculate the H2 peak during in-vessel reflooding (models for ZrO2 cracking and shattering, and clad hydriding, no reliable model for U-Zr-O mixture oxidation), the thermochemical and mechanical models in the codes underestimate the heat-up and the H2 production.
One of the main conclusions of the EU COLOSS project [20] is that oxidation of U-Zr-O and Zr-O mixtures could explain the high peak of hydrogen observed in CORA and QUENCH experiments, due to the high oxidation kinetics of these mixtures.
IRSN experts have made an extrapolation of oxidation kinetics observed in the QUENCH experiments to the 4500 MW·th EPR. This extrapolation, which is only a rough estimate, leads to reactor kinetics between 0.2 kg/s (extrapolation from QUENCH-01 results) and 7.5 kg/s (extrapolation from QUENCH-03 results). Extrapolation by IRSN for the H2
peak release during in-core reflooding of a degraded core for a French 2800 MW·th PWR is around 1 kg/s.
These extrapolations, to be used only with care for the purpose of comparison, show that the variation of hydrogen production rates during in-vessel reflood can be very large, and kinetics very different from one transient scenario to another. This effect has to be taken into account in calculations of scenarios with reflooding when checking if no possible ‘cliff edge effect’ could exist in the plant regarding the hydrogen risk.
The research on in-vessel reflooding is still ongoing. Future tests are planned in the QUENCH facility for the years to 2010.
2.2 3. In-vessel hydrogen production coming from steel oxidation
Steel oxidation may contribute about 10% to 15% of the total in-vessel hydrogen production [14]. Similar to Zr, sufficiently accurate oxidation correlations are available for an intact geometry, but no molten steel oxidation correlation exists at the present time.
2.2.4. In-vessel hydrogen production coming from B4C absorber material oxidation
During the melting of the core, when steam comes into contact with the remaining B4C inside the control rods, it is likely that any exposed B4C will react rapidly with the steam atmosphere. The B4C–steam reactions are summarized in Table 4 [25]. The energies released at 1500 K and 2200 K according to Ref. [21] are also included in this table.
The gas phase produced involves H2, CO, CO2 and CH4. In addition, vapours of B2O3 and various acids of boron are produced. It is important to add to these reactions those possible between the gases produced and the gaseous atmosphere (H2/steam ratio) of the primary circuit under severe accident conditions. If the gaseous fission products contained in the primary circuit and in the containment are neglected, these reactions are the following ones:
